Get nominal! (Never reason from a RGDP change)

One of the most disturbing aspects of modern Keynesianism is the tendency for Keynesian pundits to use RGDP data as evidence of aggregate demand changes.  And I don’t mean just occasionally, I mean 98.6% of the time.

In fact, RGDP can be affected by either demand or supply shocks.  If one uses RGDP changes as evidence of demand shocks, then the Keynesian model becomes close to tautological.  RGDP weakness must be caused by demand-side problems, because poor RGDP data is defined as a demand-side problem!  Thus you have the absurdity of people like Noah Smith proclaiming Paul Krugman as some sort of superhero, because he predicted that fiscal austerity would reduce AD in Britain.  When British RGDP does poorly this is trumpeted as “proof” that the Keynesian model is correct, even though supply-side problems explain at least 75% of the RGDP slump, and perhaps even more.

Of course there’s a much better way—use NGDP data as evidence of demand-side problems.  If Keynesians would do this they’d find lots of evidence of demand shortfalls, even in Britain.  But they also might find that the rest of us take their views more seriously.  Right now non-Keynesians simply tune out everything they say;  “Oh there they ago again, blaming all the problems of the world on demand shortfalls.”

You often hear people say “Get real.”  Keynesians need to get nominal.

PS.  On occasion a Keynesian will point to low inflation numbers as evidence of demand-side weakness.  For instance, Mike Konczal did so in this recent column.  Can someone please find all the cases where Konczal, Krugman, or any other Keynesian pointed to the unusually high inflation in Britain since 2008 as evidence excessively high AD?

PPS.  One other advantage.  Krugman complains that Lucas, Cochrane, Barro, Fama, etc, are spouting nonsense.  I guarantee that if Krugman had made his argument in NGDP terms (i.e. too little M*V) then the Chicago group would have had a much better idea of what the Keynesians were trying to argue.  As it is the two groups were just talking past each other.



12 Responses to “Get nominal! (Never reason from a RGDP change)”

  1. Gravatar of Luis Enrique Luis Enrique
    29. April 2013 at 07:56

    this paper might interest some

    Beaudry and Portier on non-inflationary demand driven business cycles

  2. Gravatar of Becky Hargrove Becky Hargrove
    29. April 2013 at 08:17

    Get nominal! Supply shocks are the good and bad decisions of policymakers, diamonds being pulled from the ground, newly discovered oil…and they all work their way into the economy at some point in highly dramatic ways which the vagaries of interest rate targeting love to illustrate and fawn over. But in the real world of nominal, it just keeps happening incrementally: those resources and decisions get implemented one step, one hour at a time, from our own collective capacity.

  3. Gravatar of JSeydl JSeydl
    29. April 2013 at 09:31

    Keynesians don’t just talk about real GDP; they more often talk about the output gap, which directly estimates the demand shortfall. You can disagree with one’s measure of potential GDP — maybe for the UK it should be lower — but assuming that it’s accurate, I don’t see what the problem is.

  4. Gravatar of ChargerCarl ChargerCarl
    29. April 2013 at 09:54

    I wouldn’t count on that Scott. Cochrane seems just as mystified by your views on his blog as he does with Krugman’s.

  5. Gravatar of Geoff Geoff
    29. April 2013 at 10:11

    Get real and nominal!

    Let’s get realominal. Let’s use the same economic laws we use for real goods, that we do with money, to create a cohesive, yet parsimonious, economic science.

    Then the “disturbing” aspects of inflationist economics can be rectified.

    Woops I forgot. Economists need to separate fields of inquiry so that they can secure little fiefdoms for themselves. I like real ice cream! Oh yeah, well I like nominal ice cream instead! It’s easier to deal with and simpler.

    Until infinity…

  6. Gravatar of RN RN
    29. April 2013 at 10:58

    I just love how Krugman gets under your skin. It makes me happy. 🙂

  7. Gravatar of Adam Adam
    29. April 2013 at 11:19

    About that PPS, you really think so? I’m highly skeptical that the disagreement was is/was about evidence and not ideology.

  8. Gravatar of Doug M Doug M
    29. April 2013 at 11:30

    GDP, real and nominal, is a measure of output and not demand. Supply can exceed demand and manifests itself as goods produced and not sold. i.e. inventory accumulation.

    Why don’t economists look at something like final sales?

  9. Gravatar of Mark A. Sadowski Mark A. Sadowski
    29. April 2013 at 11:42

    My previous comment concerning IMF projections on your “Krugman’s double standard” post was also posted at Economist’s View and a commenter there asked me to explain how I knew that Japan and the UK had experienced a positive and a negative shift in AS since 2009 respectively. (Yes, he seemed to be confused by the RGDP changes.) Here is what I wrote.

    We have to compare rates of change in NGDP, RGDP and the GDP implicit price deflator over 2009-2013 to what is “normal” for each currency area.

    The best way to identify what is normal is to find two years reasonable far apart (say a decade) at similar points in the business cycle (similar output gaps) and compute the average rates of change. Preferably the output gaps should be as close to zero as possible. (I’m using IMF data for consistency and comparability.)

    For the US and Japan this is reasonable easy. 1997/2007 works well for the US and 1995/2006 works well for Japan. For the eurozone and the UK it is less so. The eurozone is has only been around since 1999. The best years for comparison are 2000/2008 although the eurozone was nearly 2% above potential GDP in both years. The UK had a long period being above potential in the oughts and below potential in the 1990s. The best pair is 1994/2009 although the UK was nearly 2% below potential GDP in both years.

    For the given periods I find that the average rate of growth in NGDP was 5.3%, 3.9%, 0.1% and 4.7% in the US, the eurozone, Japan and the UK respectively. Growth in RGDP averaged 3.0%, 1.8%, 1.1% and 2.5% in the US, the eurozone, Japan and the UK respectively. The rate of change in the GDP implicit price deflator was 2.3%, 2.1%, (-1.0%) and 2.2% in the US, the eurozone, Japan and the UK respectively.

    Assuming no change in AS a change in AD should increase or decrease RGDP growth and inflation in the same direction. Morover, empirically the slope of the AS curve is such that a decline in NGDP growth is usually reflected roughly two thirds in real growth and one third in inflation.

    From 2009-2013 NGDP, RGDP and inflation averaged 3.8%, 2.1% and 1.7% in the US respectively, and in the eurozone it averaged 1.8%, 0.6% and 1.2% respectively. Thus the reduction in the rate of NGDP growth was reflected in about 60% real growth and 40% inflation in both the US and the eurozone. No sign of a change in AS here.

    Now, let’s look at Japan and the UK. From 2009-2013 NGDP, RGDP and inflation averaged 0.5%, 1.9% and (-1.4%) in Japan respectively and in the UK it averaged 3.0%, 0.9% and 2.1% respectively.

    Japan actually had above normal NGDP growth over this period which isn’t surprising since it had both monetary and fiscal policy moving in the same direction (unlike everybody else). But the interesting thing is that the rate of inflation decreased and the rate of real growth increased. That’s a classic indicator of a positive AS shock.

    On the other hand the reduction in the rate of NGDP growth in the UK was reflected almost entirely in a reduction in real growth. That indicates a decrease in AS.

  10. Gravatar of Rob Rawlings Rob Rawlings
    29. April 2013 at 12:51

    Falling RGDP could be due to at least 3 things

    – Supply-side issues (max output has been reduced)
    – Expectations issues (even if prices were perfectly flexible demand would be reduced)
    – Monetary issues (demand for money combined with sticky prices has artificially reduced demand).

    By making sure that NGDP stays on target you can eliminate the last factor but as the other 2 are so tightly entwined it would be hard to tell them apart.

    I’m not sure how seriously some Keynsian’s take supply-side problems. Don’t they think that if all else fails then direct govt investment in infrastructure can always be used to get RGDP to where they want it to be ?

  11. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    29. April 2013 at 13:58

    James Hamilton keeps up the pressure on Colbertmetrics;

    ‘…some researchers at the University of Massachusetts have challenged one tiny detail of one follow-up study by Reinhart and Rogoff…. That dispute concerns a measure of the correlation between sovereign debt levels and GDP growth. I say it is a tiny detail, because the dispute is completely restricted to just one of six different summaries of three different data sets in that one particular paper. The Massachusetts researchers correctly noted that there is an error in the spreadsheet which, if corrected, would have changed the number Reinhart and Rogoff should have reported for that one statistic by a few tenths of a percent. Bigger changes in that one statistic could be obtained if one adds some additional data and makes what I regard as a questionable change in methodology, but even with all the changes they want to make, the results preferred by the University of Massachusetts researchers are in fact very similar to the other 5 dataset summaries that will be found in Reinhart and Rogoff’s original paper, as well as a subsequent analysis of expanded data that Reinhart and Rogoff published in 2012 (which again the Massachusetts scholars did not mention or discuss).’

  12. Gravatar of ssumner ssumner
    29. April 2013 at 17:29

    JSeydl, You said;

    “Keynesians don’t just talk about real GDP; they more often talk about the output gap, which directly estimates the demand shortfall. You can disagree with one’s measure of potential GDP “” maybe for the UK it should be lower “” but assuming that it’s accurate, I don’t see what the problem is.”

    I’m not talking about academic papers, I’m talking about blog posts, op eds, etc. Those talk about RGDP all the time, and rarely mention NGDP. And output gaps are almost impossible to estimate reliably. What is the output gap in Britain?

    RN, I love seeing one Keynesian after another coming over here with junior high school level insults. Not one argument presented against my views. It tells me I’m winning, the other side lacks any good arguments. Even Matt Yglesias ridiculed the Keynesian argument in a post today–and he’s a highly respected progressive. If you can’t convince Yglesias, what hope to you have to be taken seriously by mainstream economists?

    Adam, It’s not about evidence or ideology, they are talking completely different languages. I did posts on this earlier–showing they were talking past each other. The paper that supposedly showed Cochrane didn’t understand that fiscal stimulus could boost AD, had Cochrane also mentioning that it could boost V, and hence NGDP. But that is AD!

    Doug, The term “demand” is misleading, it doesn’t really mean demand. We should stop talking about demand and talk about NGDP.

    Mark, That all sounds pretty plausible to me. But of course these things are very difficult to estimate.

    Rob, I don’t know, You’ll have to ask the Keynesians.

    Patrick, I love his recent posts—I’ll do a post.

Leave a Reply